bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 Biochemical Characterization of RecBCD Enzyme from An Antarctic Pseudomonas Species and

2 Identification of Its Cognate Chi (c) Sequence

3

4 Theetha L. Pavankumar*, 1,2, #, Anurag Kumar Sinha*, 1, 3 and Malay K. Ray1, #

5

6 1CSIR-Centre for Cellular and Molecular Biology, Uppal Road, Hyderabad-500007, INDIA

7 2 Department of Microbiology and Molecular Genetics, Briggs Hall, One Shields Ave, University

8 of California, Davis, CA, USA.

9 *These authors have contributed equally to this work

10 Present Address:

11 3 Department of Biology, University of Copenhagen, Ole Maaløes Vej 5, Copenhagen, Denmark.

12

13 Address for correspondence:

14 # Theetha L Pavankumar

15 Department of Microbiology and Molecular Genetics,

16 One Shields Ave, University of California, Davis, CA-95616, USA

17 Telephone: +1 530-754-9702

18 Fax: +1 530-754-8973

19 E-mail: [email protected]

20

21 # Malay K. Ray

22 CSIR-Centre for Cellular and Molecular Biology

23 Uppal Road, Hyderabad 500 007, INDIA

24 Telephone: +91 40 2719 2512

25 Fax : +91 40 2716 0591

26 E-mail : [email protected]

1 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

27 ABSTRACT

28 Pseudomonas syringae Lz4W RecBCD enzyme, RecBCDPs, is a trimeric protein complex

29 comprised of RecC, RecB, and RecD subunits. RecBCD enzyme is essential for P. syringae

30 growth at low temperature, and it protects cells from low temperature induced replication arrest.

31 In this study, we show that the RecBCDPs enzyme displays distinct biochemical behaviors. Unlike

32 E. coli RecBCD enzyme, the RecD subunit is indispensable for RecBCDPs function. The RecD

33 motor activity is essential for the Chi-like fragments production in P. syringae, highlighting a

34 distinct role for P. syringae RecD subunit in DNA repair and recombination process. Further, the

35 ssDNA-dependent endonuclease activity is notably absent in RecBCDPs enzyme. Here, we

36 demonstrate that the RecBCDPs enzyme recognizes a unique octameric DNA sequence, 5ʹ-

37 GCTGGCGC-3′ (ChiPs) that attenuates nuclease activity of the enzyme when it enters dsDNA

38 from the 3′-end. We propose that the reduced translocation activities manifested by motor-

39 defective mutants cause cold sensitivity in P. syrinage; emphasizing the importance of DNA

40 processing and recombination functions in rescuing low temperature induced replication fork

41 arrest.

42

43

44 Keywords:

45 RecBCD enzyme, Pseudomonas, Chi sequence, cold adaptation, DNA damage, DNA repair and

46 recombination, DNA helicase and nuclease.

47 Abbreviations:

48 ATP, Adenosine triphosphate; DSB, double-strand break; ‘Chi, Crossover hotspot instigator; Ni-

49 NTA, Nitrio tri-acetic acid; TLC, thin layer chromatography; MMC, mitomycin C; UV light, Ultra

50 violet; ABM, Antarctic bacterial medium; LB, Luria-Bertani medium.

2 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

51 INTRODUCTION

52 The RecBCD enzyme-mediated homologous recombination is a DNA repair pathway that

53 ensures genome integrity by faithful repair of broken DNA in E. coli and many Gram-negative

54 bacteria. The heterotrimeric RecBCD enzyme complex, comprised of RecB, RecC, and RecD

55 subunits, is essential for double-strand breaks (DSBs) repair via homologous recombination and

56 protects host cells from foreign DNAs and invading phages [1-4]. DSBs are generated in cells by

57 various exogenous and endogenous factors including the running of replication forks into

58 preexisting lesions [5]. RecBCD enzyme is a highly processive helicase and nuclease, which

59 unwinds and degrades DNA strands asymmetrically from a blunt or nearly blunt dsDNA [2].

60 Initially, RecBCD enzyme degrades 3′-ended DNA preferentially over the 5′-end of the DNA until

61 it encounters a regulatory DNA sequence called ‘Chi’ (Crossover hotspot instigator; 5′-

62 GCTGGTGG-3′ in E. coli) [2,6,7]. Chi (c) recognition switches RecBCD enzyme’s polarity of DNA

63 degradation. It attenuates 3′®5′ nuclease activity but upregulates 5′®3′ nuclease activity resulting

64 in the production of 3ʹ-ended ssDNA tail [8]. The RecBCD enzyme loads RecA onto the 3ʹ-

65 terminal ssDNA [9] facilitating the homologous pairing to form Holliday junction. The RuvAB/C

66 complex further resolve these recombination DNA intermediates to promote DNA repair process

67 via homologous recombination [10]. The behavior of RecBCD enzyme was also studied using

68 single molecule techniques [11]. Single molecule analyses of E. coli. RecBCD enzyme revealed

69 that it translocates on DNA at a much higher rate, before Chi and pauses at Chi [12]. The Chi

70 recognition switches lead motor subunit of the RecBCD enzyme, from fast to slow motor (RecD

71 to RecB), resulting in the reduction of translocation rate by one-half after Chi [13].

72 Previously, we have shown that the RecBCD enzyme of Antarctic Pseudomonas syringae

73 Lz4W (RecBCDPs) is essential for the growth at low temperature [14]. Growth at low temperature

74 induces frequent replication arrest causing fork reversal mediated DNA damage in P. syringae

75 Lz4W [15]. The RecBCDPs enzyme thus rescues cells from replication-arrest dependent DNA

3 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

76 damage enabling P. syringae Lz4W to grow at low temperature [15]. Unlike in E. coli, the RecD

77 subunit of P. syringae Lz4W is an indispensable subunit of RecBCD complex [16]. The genetic

78 analysis of ATP binding and nuclease defective mutants of RecBCDPs enzyme indicated that the

79 ATP driven motor activities of both RecD and RecB motor subunits are critical for growth at low

80 temperature, whereas the nuclease activity of holoenzyme is dispensable [14].

81 In this study, we performed a biochemical analysis of wild-type and mutant RecBCDPs

82 enzymes to understand their biochemical role in protecting P. syringae Lz4W cells from low

83 temperature induced DNA damage. Here, we report that the ATP dependent DNA unwinding, not

84 the DNA degradation activity of RecBCD enzyme is critical for growth at low temperature.

85 Besides, inactivation of ATPase activity of RecD and RecB motor subunits has impaired the DNA

86 unwinding activity of RecBCD enzyme leading to cold-sensitive phenotype in vivo. Furthermore,

87 we have identified the P. syringae Chi sequence (5′-GCTGGCGC-3′), which attenuates the

88 RecBCDPs nuclease activity in vitro.

89

90

4 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

91 RESULTS

92 Purification of RecBCD complex and its mutant variants from P. syringae Lz4W

93 The P. syringae recCBD genes were overexpressed on pGL10 derived plasmids in

94 ∆recCBD (LCBD) strain to obtain RecB, RecC (N-terminal His-tagged) and, RecD proteins in an

95 equimolar ratio (13). The wild-type and mutant RecBCD enzymes (RecBK28QCD, RecBCDK229Q,

96 and RecBD1118ACD) were purified by two-step purification protocol as shown in the flowchart (Fig.

97 1A) using Ni-NTA affinity column chromatography and size exclusion chromatography. The

98 purified fractions of RecBCD and mutants contained all the three protein subunits (RecB, RecC,

99 and RecD) in a stoichiometric ratio (Fig. 1B). Silver nitrate staining of gels further confirmed the

100 purity of protein fractions. However, silver staining of SDS-PAGE separated proteins showed an

101 additional protein band of ~60 kDa size (Fig. 1C). The Mass spectrometric analysis of this protein

102 band indicated that it belongs to HSP-60 family of chaperonin, GroEL. The appearance of GroEL

103 as a contaminant during purification of RecBCD protein is also evidenced in E. coli [17].

104 Nonetheless, the presence of RecB, RecC, and RecD subunits was further confirmed by Western

105 analysis using the polyclonal antibodies specific to these proteins (Fig. S1).

106 ATP hydrolysis activity of the wild-type and the mutant RecBCD enzymes

107 ATP hydrolyzing activity of wild-type and mutant RecBCD enzymes was measured by thin

108 layer chromatography (TLC) as described in Materials and methods. The wild-type RecBCD

109 enzyme displayed DNA stimulated ATPase activity in the presence of linear pBR322 double-

110 stranded DNA (dsDNA). At 22ºC, RecBCD hydrolyzed ATP with the maximum velocity (Vmax) of

-1 111 236.5 µmol ATP per µmol enzyme s and a Km of 57.8 µM for ATP (Fig. 2A and Table 1). The

112 mutant RecBD1118ACD enzyme (nuclease defective mutant) showed ATPase activity similar to the

113 wild-type enzyme. However, the ATP-binding defective mutant enzymes such as RecBK28QCD

114 (mutation in the consensus ATP binding site of RecB) and RecBCDK229Q (mutation in the

115 consensus ATP binding site of RecD) showed a 10-fold decrease in ATP hydrolyzing activities

116 (Table 1).

5 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

117 ATP hydrolyzing activities of RecBCD enzymes were measured at three different

118 temperatures (37, 22 and 4°C) (Fig. 2B). Interestingly, The RecBCD enzyme showed the highest

119 ATPase activity at 37°C compared to 22°C and 4°C. At 22°C, the optimum temperature for P.

120 syringae Lz4W growth, the wild-type RecBCD and the nuclease-deficient RecBD1118ACD enzymes

121 displayed ~40% lower ATPase activity compared to 37°C. At 4°C, the activities were further

122 reduced to ~50% of their respective ATPase activity observed at 22°C. The ATP hydrolyzing

123 defective mutants showed ~10-fold decrease in ATP hydrolyzing activity compared to wild-type

124 RecBCD enzyme, at all the temperatures tested (Table 1 and Fig. 2C). It indicates that the

125 mutations were previously shown to prevent P. syringae growth at low temperature severely

126 curtailed ATP hydrolyzing activity.

127 DNA unwinding and degradation activities of the RecBCD enzyme at different

128 concentration of Mg++ and ATP

129 The DNA unwinding and degradation activity of E. coli RecBCD enzyme (RecBCDEc) are

130 free [Mg2+] ion dependent. An increase in free [Mg2+] increases nucleolytic cleavage by RecBCD

131 enzyme [18]. To understand the DNA unwinding and degradation behavior of P. syringae

132 RecBCD enzyme (RecBCDPs), we performed experiments at various concentrations of

133 magnesium and ATP as described in Materials and methods. In the first set of reactions, the molar

134 concentration of magnesium was kept constant (2 mM) and the ATP concentration was varied (0-

135 10 mM); and in the second set of reactions, the ATP concentration was kept constant (2 mM) and

136 magnesium concentration was varied (0-10 mM). The results indicate that the DNA unwinding

137 and degradation properties of RecBCDPs enzyme are also modulated by the ratio of Mg2+ and

138 ATP. When the molar concentration of Mg2+ is lesser than ATP, RecBCDPs enzyme unwinds the

139 dsDNA substrate, and the degradation of DNA is not much pronounced (Fig. 3A, lane 4-6).

140 However, when the molar concentration of Mg++ exceeds the ATP concentration, RecBCDPs

141 degrades unwound DNA more vigorously (Fig. 3B, lane 4-10). The ratio of [Mg2+]: [ATP] thus

6 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

142 affects the kinetics of DNA unwinding and degradation properties of RecBCD enzyme. We also

143 observed that an increase in ATP concentration more than three folds over Mg++ inhibits the DNA

144 unwinding activity of RecBCDPs enzyme (Fig. 3A, lane 7-10). The inhibition of DNA unwinding

145 activity is possibly due to sequestration of a Mg2+ ion by ATP leading to the depletion of free

146 magnesium in the reaction. Subsequently, the DNA unwinding and degradation experiments were

147 performed under specified reaction conditions. The limiting magnesium reaction condition (5 mM

148 ATP, 2 mM Mg++) was chosen to observe the DNA unwinding activity and, the excess magnesium

149 condition (2 mM ATP, 6 mM Mg++) to observe the DNA unwinding and degradation activities of

150 RecBCD enzyme. Interestingly, under excess magnesium conditions, we observed three shorter

151 discrete DNA fragments (Fig. 3B, Lane 5-10). These DNA fragments production by RecBCD

152 enzyme is possibly due to Chi-like sequence on a DNA substrate and is discussed later in the

153 results section.

154 Calcium has been shown to inhibit the nuclease activity of E. coli RecBCD enzyme [19].

155 We also studied the effects of calcium on P. syringae RecBCD by increasing the Ca++ ion

156 concentration in a reaction mixture that contained fixed amounts of ATP and Mg2+ (2 and 6 mM

157 respectively) (Supplementary Fig. S2). We found that similar to RecBCDEc, the high concentration

158 of calcium inhibits both helicase and nuclease activity of RecBCDPs enzyme.

159 Temperature-dependent DNA unwinding and nuclease activity of P. syringae RecBCD

160 enzyme

161 ATPase RecBCD mutants affect P. syringae Lz4W growth in a temperature-dependent

162 manner. Hence, we sought to assess the effects of temperature on the DNA unwinding and

163 degradation properties of RecBCDPs enzyme. We measured the enzyme activities at 22°C, and

164 4°C, using NdeI linearized [5′-32P] labeled pBR322 plasmid as a substrate. Under the limiting

165 magnesium reaction condition (5 mM ATP, 2 mM Mg2+), the DNA unwinding activity (i.e.,

166 production of full-length ssDNA) of RecBCDPs enzyme was detected only at 22°C and not at 4°C

7 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

167 (Fig. 4A). However, under excess magnesium reaction condition (2 mM ATP, 6 mM Mg++) the

168 DNA degradation activity of RecBCDPs enzyme was observed at both 22°C and 4°C (Fig. 4B).

169 We calculated the rate of DNA unwinding by measuring the decreased band intensities of

170 5′-[g-32P]-labeled dsDNA substrates at different temperatures. The wild-type RecBCDPs unwound

171 the DNA at the rate of 35.1 ± 1.6 bp/sec at 22°C when ATP and Mg++ ratio was 5:2 (limiting

172 magnesium condition) (Table 2). However, the enzyme could unwind the DNA even much faster,

173 when the ATP and Mg++ ratio was changed to 2:6 (excess magnesium condition). Under the latter

174 condition, wild-type RecBCDPs enzyme unwound (also degraded) the DNA at the rate of 101.5 ±

175 3.3 bp/sec (Table 2). Notably, under the limiting magnesium reaction condition at low temperature

176 (4°C), the RecBCD enzyme failed to produce detectable unwound DNA products (Fig. 4A).

177 However, under excess magnesium conditions, RecBCD enzyme could unwind and degrade the

178 DNA at the rate of 55.8 ± 6.9 bp/s (Table 2, Fig. 4B), which is about 54% of the activity observed

179 at 22°C. The apparent unwinding rates obtained under excess magnesium conditions, based on

180 the disappearance of 32P-end-labeled DNA substrate could be an over-estimation. It is possible

181 that some enzyme molecules could just nucleolytically remove the end-labeled nucleotide, but

182 couldn’t fully unwind the dsDNA substrate.

183 DNA unwinding and nuclease activities of mutant RecBK28QCD, RecBCDK229Q, and

184 RecBD1118ACD enzymes

185 We have shown that P. syringae cells carrying recBK28QCD or recBCDK229Q mutants are

186 sensitive to cold temperature, UV irradiation and Mitomycin C (MMC) [14]. These two mutant

187 enzymes also display very weak ATPase activity. To understand the impact of these mutations

188 on the DNA unwinding and degradation properties of RecBCD complex, we analyzed RecBK28QCD

189 and RecBCDK229Q enzymes in vitro at 22°C and 4°C. At 22°C, under limiting and excess

190 magnesium conditions, RecBK28QCD with the ATPase-defective RecB subunit displayed weak

191 helicase activity (Fig. 5A). The rate of DNA unwinding was 1.5 ± 0.7 bp/sec and 12.9 ± 3.4 bp/sec

8 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

192 at 22°C, under limiting and excess magnesium conditions respectively (Fig.5A and Table 2). At

193 4°C, on the other hand, RecBK28QCD showed no detectable DNA unwinding activity under the

194 excess magnesium condition (Fig. 5A and Table 2). These results suggest that RecBK28QCD is a

195 poor helicase-nuclease enzyme.

196 Compared to the RecBK28QCD enzyme, RecBCDK229Q enzyme with the ATPase-defective

197 motor RecD subunit displayed higher DNA unwinding and nuclease activities. At 22°C, the

198 RecBCDK229Q enzyme with the ATPase-defective RecD subunit displayed substantial DNA

199 unwinding activity, 27.2 ± 6.1 bp/sec and 92.6 ± 14.1 bp/sec, under limiting and excess

200 magnesium conditions respectively (Fig. 5B and Table 2). However, at 4°C under excess

201 magnesium conditions, the mutant enzyme unwound the DNA at the rate of 17.3 ± 9.6 bp/sec (Fig

202 5B and Table 2). The combined DNA unwinding and degradation activity of the RecBCDK229Q

203 enzyme (under excess magnesium) were ~90% and ~30% of the wild-type RecBCDPs activity at

204 22°C and 4°C respectively. Interestingly, the mutant RecBCDK229Q enzyme failed to produce

205 discrete DNA fragments (Fig. 5B).

206 We also tested DNA unwinding and degradation by the nuclease defective RecBD1118ACD

207 enzyme. (Fig. 5C). The mutation in the nuclease catalytic site of RecB subunit (RecBD1118ACD)

208 led to a complete loss of in vivo nuclease activity in RecBCDPs complex, without affecting the

209 recombination proficiency and cold adaptation of bacteria [14,15]. At 22oC, under limiting

210 magnesium reaction condition, RecBD1118ACD enzyme produced single-stranded pBR322 DNA at

211 the rate 30.4 ± 1.7 bp/sec, which is 85% of the rate observed with the wild-type enzyme. Under

212 excess magnesium condition, the rate of DNA unwinding increased to 109.7 ± 17.5 bp/sec, which

213 is similar to wild-type. At 4°C, this enzyme could unwind the pBR322 DNA at the rate of 41.0 ±

214 8.1 bp/sec, which is about 75% of wild-type enzyme (Table 2). More importantly, under both

215 limiting and excess magnesium conditions, this enzyme as expected, produced only the full-length

216 ssDNA of pBR322 and did not degrade the DNA (Fig. 5C). From these data, it is clear that

9 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

217 RecBD1118ACD enzyme is nuclease deficient in vitro, but its DNA unwinding activity is comparable

218 to wild-type enzyme at both 22 and 4°C (Fig. 5C and Table 2).

219 P. syringae RecBCD enzyme does not have endonuclease activity

220 The ATP-dependent endonuclease activity of wild-type RecBCD enzyme of P. syringae

221 was examined using a circular M13 ssDNA as substrate as previously described for RecBCDEc

222 [20-22]. We tested the endonuclease activity under three different conditions of ATP and

223 magnesium concentrations as described in Supplementary Fig S4. Under all the three conditions,

224 wild-type RecBCD and mutant proteins failed to degrade M13 phage ssDNA (Fig S3). It indicates

225 that unlike RecBCDEc, the RecBCDPs does not exhibit endonucleolytic activity under the

226 conditions tested.

227 A specific DNA sequence on pBR322 plasmid modulates nuclease activity of P. syringae

228 RecBCD.

229 We noticed that, under the excess magnesium reaction conditions, RecBCDPs produced

230 a discrete DNA fragments shorter than the full-length unwound DNA substrate (Fig 3B, Lanes 5-

231 10), and the shortest DNA fragment showed a higher intensity than the other ones (Fig 3B). This

232 interesting observation led us to hypothesize that the P. syringae RecBCD enzyme recognizes a

233 specific DNA sequence. This specific DNA sequence could potentially be a Chi (c) like sequence;

234 which alters the nuclease activity of RecBCDPs complex, allowing 3ʹ-ended ssDNA to escape from

235 DNA degradation as observed earlier [2,8,23,24]. To further confirm that these DNA fragments

236 are indeed single-stranded, we performed degradation assays in the presence of Exonuclease-I

237 (ExoI), which specifically degrades ssDNA by its 3′®5′ exonuclease activity [25]. The RecBCDPs-

238 generated short DNA fragments disappeared in the presence of ExoI (Fig S4A) suggesting that

239 they are indeed ssDNA, similar to Chi-specific DNA fragments produced by E. coli RecBCD

240 enzyme after Chi-recognition [8,23,24,26].

10 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

241 Using genetic experiments, we have previously shown that RecBCDPs enzyme does not

242 recognize E. coli Chi sequence (5ʹ-GCTGGTGG-3') [14,16]. We performed biochemical

243 experiments using pBR322c+3F3H (a pBR322 derivative, containing three tandem E. coli c

244 sequences) [27] and pBR322 (lacking E. coli c sequence) as a DNA substrate. Interestingly,

245 ssDNA fragments produced by RecBCDPs enzyme were identical in size with both the substrates,

246 confirming that RecBCDPs does not recognize E. coli Chi sequence but recognizes an unknown

247 DNA sequence of pBR322 plasmid (Fig S4B). To locate putative Chi sequence of RecBCDPs

248 enzyme, we amplified 3.6 kb segment of pBR322 using primer OROPI and OROPII

249 (Supplementary Table ST2), which excludes rop region of the plasmid (Fig. S5A). Amplified

250 fragments were 5ʹ-labeled, and the assays were performed in the presence of excess magnesium.

251 Apparently, all the three shorter ssDNA fragments were also observed with the 3.6 kb DNA

252 substrate, indicating the presence of putative Chi sequence within 3.6 kb of the plasmid

253 (Supplementary Fig. S5C).

254 Chi dependent protection of pBR322 DNA fragments are strand specific

255 E. coli RecBCD enzyme recognizes Chi sequence in a specific orientation, 5ʹ-

256 GCTGGTGG-3ʹ, when enzyme enters through 3′-end [28] and the 3ʹ-ended ssDNA is being

257 protected from RecBCD nuclease activity after Chi recognition [8,29,30]. Here, we examined

258 which strand of the linearized pBR322 is being protected to produce these discrete DNA bands

259 (Fig. 3B). Hence, we performed DNA degradation assay with RecBCDPs, in which only one DNA

260 strand was labeled at a time. We individually labeled the OROPI and OROPII primers with g32P

261 at 5′-end and amplified the 3.6 kbp region of pBR322 plasmid with one unlabeled and another

262 32P-labeled primer or with both labeled primers. In our assay, we always used sub-saturating

263 concentration of RecBCDPs enzyme compared to DNA substrate, so that RecBCDPs enzyme

264 enters through either one or the other end, but not through both the ends. DNA degradation

265 assays using these DNA substrates produced all the three bands when top strand (OROPII end)

11 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

266 was alone labeled, or when the both strands were end-labeled (Fig. 6A). We did not observe any

267 ssDNA band when the bottom strand (OROPI end) was labeled (Fig. 6A), suggesting that

268 RecBCDPs recognizes a Chi-like sequence in a specific orientation and has a polarity for Chi

269 sequence recognition. These results also suggest that all the Chi-like sequences are on the top

270 strand.

271 We then performed DNA degradation assays using two other DNA substrates amplified

272 from the pBR322 plasmid. A PCR amplified 2.8 kbp DNA fragment (using OROPII and pBRS1R

273 primers) and a 2.5 kbp PCR amplified fragment (using OROPII and pBRB1R primers set (Fig.

274 S6A). The DNA band of the lowest size (~2.4 kb) and the highest intensity (the prominent DNA

275 band) was the common ssDNA fragment produced by RecBCDPs in all the three DNA substrates

276 (Supplementary Fig.S6B). This suggests that one of the Chi-like sequences has the strongest

277 RecBCD-inhibitory activity and this site (we designate it as ChiPs) is located proximal to the 2.5

278 kbp substrate end (as the protected ssDNA was about ~ 2.4 kb). Two upper DNA bands with less

279 intensity could be due to variants of ChiPs, which might have a weak inhibitory effect on RecBCDPs

280 nuclease activity (see below).

281 Identification of ChiPs as an 8-mer (5ʹ-GCTGGCGC-3ʹ) sequence that modulates P. syringae

282 RecBCD nuclease activity and protects DNA from further degradation

283 To identify the precise location of ChiPs site in pBR322 DNA substrate, we generated

284 several internally deleted constructs of pBR322 plasmid by site-directed segment-deletion as

285 described in Materials and methods. The 3.6 kbp DNA region of pBR322 plasmid with deleted

286 region/s were PCR amplified using OROPI and OROPII primers. DNA degradation assays were

287 then performed with RecBCDPs enzyme using the DNA fragments with internal deletion as

288 substrates (Supplementary Fig. S7).

289 The localization of ChiPs on the plasmid was first based on two hypotheses: the sequence

290 might be GC rich and should be located close to the right end of 2.5 kbp DNA fragment. From the

291 pBR322 nucleotide sequence analysis, it appeared that a GC-rich region spanning the region

12 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

292 between 273-450 nucleotides within tetR gene of pBR322 might contain the ChiPs. Accordingly,

293 3.6 kbp DNA containing two deletions (∆273-381 and ∆400-450) were initially tested by the Chi-

294 protection assays.

295 The 3.6 kbp DNA deleted for 273-381 region produced the Chi-specific high-intensity

296 fragment (Supplementary Fig. S7), while 400-450 nucleotides deletion failed to produce it (Fig.

297 6B). This indicated that the ChiPs sequence is located within 400-450 nucleotide region of the

298 pBR322 plasmid. The three additional deletions within the 400-450 region of the plasmid segment

299 were made; ∆401-419, ∆421-439, and ∆441-449 (Fig. 6C, D). DNA degradation assays with the

300 fragments in which these sequences were deleted revealed that Chi is located within the 421-439

301 nucleotides segment, as these deletions did not produce a protected prominent DNA fragment

302 (Fig. 6C). we further made two more deletion constructs (∆421-429 and ∆431-439) within the 421-

303 439 nucleotides region, and tested for Chi-specific fragment production. Interestingly, deletion of

304 base-pair from 431-439 bp abolished the prominent DNA band (Fig. 6D) indicating that the deleted

305 region between 431-439 bp, 5′-GCTGGCGCC-3′ contains the putative Chi sequence of P.

306 syringae. The schematic representation of all the constructs with deleted nucleotide regions and,

307 the presence or absence of protected prominent DNA band (putative Chi sequence) are also

308 shown in Fig. 6E.

309 The identification of putative ChiPs sequence further prompted us to investigate the reason

310 for producing an apparent three distinct DNA fragments by RecBCDPs enzyme (Fig. 3B). We

311 speculated that pBR322 DNA substrate might have sequences that are similar but not identical

Ps Ps 312 to Chi sequence, and might act like Chi sequence with weaker recognition property under the

313 in vitro assay conditions. The 9-mer (5′-GCTGGCGCC-3′) sequence appears only at one place

314 on the pBR322 substrate. Considering the E. coli Chi sequence is an octamer, we first looked in

315 to 5¢-GCTGGCGC -3¢ (first 8 nucleotides of the 9-mer from the 5¢-end), and 5′-CTGGCGCC-3¢

316 (eliminating first G of the 5¢-end) as possible Chi-like sequences. These two octamer sequences

13 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

317 appear at a single location on the pBR322 substrate. However, among the 7-mer combinations

318 we looked into, the first 7 nucleotide sequence, 5¢-GCTGGGC -3¢, was located in three reasonable

319 positions on pBR322 substrate which could potentially produce discrete DNA bands as observed

320 in DNA degradation assays. Hence, we further focused on 5¢-GCTGGCGC -3¢ sequence as a

321 putative ChiPs sequence. We found that the putative ChiPs sequence (5¢-GCTGGCGC -3¢ ) is

322 located at 431-438 bp position of pBR322 and, the similar 7-mer sequences at three different

323 locations. The first one at 964-971 position (5′-GCTGGCGT-3′); the second one at 1493-1500

324 position (5′-GCTGGCGG-3′) and the third one at 2525-2532 position (5′-GCTGGCGT-3′) (Fig S8).

325 All the three-bands show 7 bases identical to the 8 bp ChiPs sequence (5′-GCTGGCGC-3′), and

326 occur in the same orientation (5′→3′) as the ChiPs. The RecBCDPs enzyme might recognize 8 mer

327 sequences (5′-GCTGGCGC-3′) as well as the 7+1 mer sequences (5′-GCTGGCG+T/G-3′)

328 resulting in multiple DNA bands. Although it is expected to observe four discrete DNA bands, we

329 observed only three DNA bands. The fourth-expected DNA band with a size of ~230 bp (when

330 RecBCDPs enzyme enters from 3’-side of the NdeI linearized pBR322 substrate as depicted in

331 Fig. S8) was not apparent in in-vitro experiments performed on agarose gels. Possibly, it is the

332 last Chi-like sequence to recognize by RecBCD enzyme and also being the shortest one with ~

333 230 bp (Fig. S8).

334 To better define the ChiPs sequence, we mutated the T to C at the 971st position of pBR322,

335 which creates a sequence identical to the putative 8-mer ChiPs sequence and performed the Chi-

336 protection assay with RecBCDPs enzyme. Interestingly, the intensity of the Chi-protected second

337 DNA fragment (after T to C mutation) appeared stronger, and the intensity was similar to the

338 prominent DNA band (Fig. 7A), which establish the strong recognition of octamer sequence (5′-

339 GCTGGCGC-3′) as ChiPs.

340 Cloning of ChiPs in pBKS plasmid and confirmation of ectopic ChiPs activity

14 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

341 To further confirm that the octamer sequence (5′-GCTGGCGC-3′) is indeed the Chi

342 sequence in P. syringae and can function as an ectopic ChiPs sequence, we cloned the

343 octanucleotide sequence into pBKS which is devoid of ChiPs sequence (Table ST1). RecBCDPs

344 reactions under excess magnesium were performed with XbaI digested linear pBKS (Chi0)

345 plasmid and the ChiPs containing pBKS (pBKS-ChiPs) plasmid. As expected, no Chi-protected DNA

346 bands were observed using the linear pBKS (Chi0) DNA as substrate. In contrast, the expected

347 Chi specific DNA fragment was observed in the case of XbaI digested pBKS-Chi plasmid (Fig.

348 7B). These experiments confirm that RecBCDPs recognizes an 8-mer sequence (5′-GCTGGCGC-

349 3′) as ChiPs sequence during resection of double-stranded DNA ends, and thus ChiPs attenuates

350 the nuclease activity of RecBCD enzyme and promotes DNA recombination and repair.

351

352

15 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

353 DISCUSSION

354 We have earlier shown that RecBCD protein complex is essential for cold adaptation in

355 Antarctic P. syringae Lz4W [14-16]. In this study, we have analyzed the biochemical properties of

356 the wild-type (RecBCDPs) and three mutant enzymes (RecBK28QCD, RecBCDK229Q, and

357 RecBD1118ACD) to analyze the role of RecBCD during growth at low temperature. We also report

358 for the first time, Pseudomonas octameric Chi-sequence (ChiPs), (5′-GCTGGCGC-3′), and its

359 ability to attenuate the nuclease activity of P. syringae RecBCD enzyme in vitro.

360 Role of ATPase activity of RecB and RecD subunits

361 The analyses of wild-type and mutant enzymes of P. syringae have revealed that mutation

362 in the critical ATP binding sites of RecB (RecBK28Q) and RecD (RecDK229Q) motor subunit reduces

363 the ATPase activity of trimeric RecBCD complex by ~10 fold. The mutation in the nuclease active-

364 site does not affect the ATPase activity. P. syringae cells carrying the alleles of RecBK28Q or

365 RecDK229Q are sensitive to cold temperature, UV irradiation and MMC [14]. The ~ 10-fold reduction

366 of ATPase activity observed in-vitro supports the idea that ATP hydrolysis by RecB and RecD

367 subunits in RecBCDPs holoenzyme is essential for P. syringae survival at low temperature.

368 DNA unwinding and degradation activities of RecBCD and mutant enzymes

369 Our data suggest that the DNA unwinding and degradation properties of RecBCD enzyme

370 depend on ATP and Mg++ concentrations, and on the temperature in vitro. Under limiting

371 magnesium reaction condition, at 22°C, the RecBK28QCD, RecBCDK229Q and RecBD1118ACD

372 enzymes have retained about 4%, 77% and 85% of the wild-type DNA unwinding activity

373 respectively, while at 4°C, we could not detect the DNA unwinding activity of both wild-type and

374 mutant RecBCDPs enzymes under the limiting magnesium reaction condition. In contrast, when

375 magnesium was in excess over ATP, the DNA unwinding and degradation activity of the RecBCD

376 enzymes could be measured 22°C and 4°C). These results suggest that the excess of magnesium

377 over ATP is favorable for the helicase and nuclease activities of RecBCDPs enzyme.

16 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

378 The RecBK28QCD enzyme is a poor helicase and with no detectable nuclease activity in

379 vitro even at 22°C. However, at 4°C, RecBK28QCD enzyme failed to unwind and degrade the DNA

380 under both limiting and excess magnesium reaction conditions. Hence, low-temperature growth

381 sensitivity of the P. syringae recBK28QCD mutant can be attributed to the lack of RecBCD

382 unwinding and/degradation activities at 4°C.

383 The RecBCDK229Q enzyme with defective RecD ATPase unwinds and degrades the

384 dsDNA, albeit at the reduced rate. The DNA unwinding/degradation rate of RecBCDK229Q mutant

385 enzyme is reduced to one-third of the wild-type enzyme at 4°C. Therefore, we propose that its

386 inability to support the DNA repair process and growth at low temperature is possibly due to its

387 decreased DNA unwinding/degradation activity, particularly at low temperature. Interestingly, the

388 RecBCDK229Q enzyme shows lack of discrete DNA fragments (putative Chi-specific fragments)

389 production. However, the E. coli RecD ATPase defective RecBCD enzyme produces the Chi-

390 specific fragments in vitro and confers DNA repair proficiency in vivo [31]. This converse

391 observation suggests that the RecD subunit of RecBCDPs enzyme has a distinct role in DNA repair

392 and recombination function in P. syringae. Despite both RecBCDK229Q and RecBK28QCD mutant

393 enzymes being similarly defective in ATP hydrolyzing activity, their DNA unwinding activity was

394 largely varied. This possibly be due to the selective motor dependency of the enzyme complex,

395 similar to observation made in E. coli RecBCD enzyme [31], in which, the RecB motor is the

396 absolute requirement for Chi recognition and the motor activity of RecBCD complex.

397 The RecBD1118ACD enzyme is a processive helicase with no detectable nuclease activity

398 in vitro, at both 22°C and 4°C. Interestingly, P. syringae cells carrying recBD1118A allele are capable

399 of growing at low temperature, suggesting nuclease activity is not essential for its growth at low

400 temperature [14].

401 RecF pathway role in RecBCD nuclease defective strain and its importance in P.

402 syringae growth at low temperature

17 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

403 In E. coli, the RecF pathway along with RecJ (5ʹ®3ʹ ssDNA specific nuclease) is known

404 to work with nuclease-defective, RecA loading-deficient RecBCD [32]. The role of RecBCD

405 nuclease defective - RecF hybrid pathway in P. syringae is also evidenced. Deletion of recF gene

406 alone (in WT cells) is cold resistant [15]. However, deletion of the recF gene in recBD1118A mutant

407 renders cold sensitivity (unpublished observation, Apuratha T Pandiyan and Malay K Ray). The

408 over-expression of RecJ in pRecBDnucCD (RecB nuclease domain deleted) expressing P. syringae

409 recBCD null strain alleviated the slow growth phenotype of P. syringae cells at low temperature

410 [14], suggesting RecJ role in nuclease-defective RecBCD cells. These observations indicate

411 RecFOR-RecJ role in RecB nuclease-defective P. syringae strain.

412 recA deleted P. syringae cells grow slowly at low temperature [15]. This suggests that

413 RecBCD enzyme alone, in the absence of RecA, can rescue low temperature-induced replication

414 fork arrest possibly by suppressing chromosomal lesions via DNA degradation of reversed

415 replication fork [33]. Interestingly, the combination of a recA deletion and a nuclease defective

416 RecBCD mutation (recBD1118ACD) causes cold sensitivity [15], suggesting a direct role of RecA in

417 rescuing low temperature induced replication fork arrest in a RecB nuclease defective strain.

418 Therefore, we propose that, in RecB nuclease defective strain, the RecF pathway enables RecA

419 mediated DNA repair and thus, protects cells from low temperature induced DNA damage.

420 Identification and characterization of Novel octameric ChiPs sequence

421 One of the novel findings in this study is the identification of P. syringae Lz4W Chi-

422 sequence (ChiPs), 5′-GCTGGCGC-3′. This sequence has not been identified so far in any

423 Pseudomonas species. Most of the identified bacterial Chi (c) sequences are GC rich sequences

424 (Table 3) and the number of nucleotides in c sites vary from 5-mer (in Bacillus subtilis) to 8-mer

425 (in E. coli, L. lactis and H. influenza) [34]. A study using cell lysate from Pseudomonads indicated

426 that Pseudomonas species do not recognize E. coli c sequence [24]. We have confirmed this

427 observation earlier by genetic experiments [14], and in the present study, we have biochemically

18 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

428 identified the ChiPs (5′-GCTGGCGC-3′) sequence; which is identical up to 5 bases from the 5′-

429 end to the E. coli Chi (ChiEc) (5′-GCTGGTGG-3′) sequence. 7-mer (5′-GCTGGCG-3′) sequence,

430 with change in the last 8th nucleotide, are partially recognized. The mutation of 7-mer sequence

431 to make a perfect 8-mer ChiPs sequence enables it to be strongly recognized by RecBCDPs.

432 Interestingly, appearance of three ssDNA fragments indicates that RecBCDPs enzyme can

433 sometimes bypass ChiPs and recognize the next putative Chi-like sequence. Similar observations

434 were made in E. coli, where the probability of Chi recognition by E. coli RecBCD enzyme and

435 nicking the DNA is about 30-40% [28,35].

436 Interestingly, the difference between E. coli and P. syringae RecBCD enzymes are

437 confined to the last three nucleotides (5¢- GCTGGTGG -3¢ vs 5¢- GCTGGCGC -3¢). The recent

438 study on the molecular determinants responsible for the Chi recognition by RecBCD enzyme has

439 revealed the importance of RecC channel in Chi recognition. Among the 35 amino acid residues

440 of RecC channel examined, the Q38, T40, Q44, L64 W70 D133, L134, D136, D137, R142, R186

441 and D705 residues of E. coli RecC subunit have shown to affect the Chi recognition property of

442 RecBCDEc enzyme [36]. Surprisingly, all these residues are well conserved in the P. syringae

443 RecC subunit. Therefore, the RecC amino acids responsible for recognition of the last three

444 nucleotides of are still elusive. Further analysis of E. coli and P. syringae RecC subunits could

445 shed more insight on the molecular determinants responsible for the recognition of last three

446 nucleotides of Chi sequence in E. coli.

447 E. coli contains 1,008 Chi sequences [37]. They are four- to eightfold more frequent than

448 expected by chance and appear on average once every 4.5 kbp. 75% of Chi sites are skewed

449 towards the replicative leading strand in E. coli [37] keeping with their function in stimulating

450 double-strand break repair upon replication fork collapse. These observations suggest a role for

451 the RecBCD enzyme as a repair factor functioning towards re-establishment of DNA replication

452 fork in case of collapse. However, this skewed nature is not applicable for B. subtilis, S. aureus

453 and H. influenzae, where the skew is statistically insignificant, and the Chi sequence (of the

19 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

454 respective species) is distributed all over the genome [34]. The search for ChiPs sequence (5′-

455 GCTGGCGC-3′) in the draft genome sequence of P. syringae Lz4W (4.98 Mb in 42 contigs,

456 Accession no. AOGS01000000) revealed that it contains 1541 ChiPs sequences (5′-

457 GCTGGCGC-3′) and 4564 of 7-mer ChiPs sequences (5′-GCTGGCG-3′). The ChiPs sequence

458 appears once in every ~3 kb and is overrepresented compared to other random octamer

459 sequences. No skewed ChiPs sequence distribution was observed in P. syringae Lz4W genome

460 (A. Pandiyan and M. K. Ray, unpublished observation). Analysis of the closely related P.

461 fluorescens Pf0-1 genome (Accession no. NC_007492.2) [38] revealed that it contains 2241

462 ChiPs sequences and the ChiPs sequence appears once in every 3 kb and, is almost equally

463 distributed on both strands (1119 Chi-Ps vs. 1122 complementary Chi-Ps sequences). Thus the

464 pattern of Chi-distribution is not universal, and although ChiPs is over-represented in

465 Pseudomonas genome, the orientation bias is not observed.

466 Biochemical properties of RecBCD enzyme and its role in P. syringae growth at low

467 temperature

468 This study has revealed the biochemical properties of Pseudomonas RecBCD enzyme.

469 The biochemical properties of RecBCDPs enzyme, compared to the RecBCDEc, are particularly

470 associated with the RecD subunit. In E. coli, RecD is dispensable for DNA repair process. The

471 recD null E. coli strain is hyper-recombinogenic [39] and RecBCEc enzyme (without RecD)

472 unwinds dsDNA and loads RecA constitutively in a Chi-independent manner [30]. Also, the E. coli

473 RecBCD enzyme with a mutation in the ATP binding site of RecD subunit produces Chi-specific

474 fragments and cells expressing the mutant enzyme are UV resistant [31]. In contrast, P. syringae,

475 RecD is essential for the RecBCD’s function [14] and ATP hydrolyzing activity of RecD motor is

476 an absolute requirement for ChiPs fragments production. Also, cells expressing RecD ATPase

477 mutant enzyme are cold sensitive, UV and MMC sensitive (13). Importantly, Chi-like octameric

478 sequence (5′-GCTGGCGC-3′) attenuates nuclease activity of RecBCDPs enzyme producing ChiPs

20 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

479 containing ssDNA fragments and thus, can act as a Chi sequence for RecBCD enzyme of

480 Pseudomonas species.

481 Based on our results we propose a model (Figure 8) which explains the role of RecBCD

482 and collaborative DNA repair pathways in rescuing the replication fork arrests in P. syringae Lz4W

483 at low temperature. In this model, RecBCD enzyme can rescue replication fork from the arrest by

484 linear chromosomal DNA degradation in a recA-independent manner. When nuclease activity is

485 compromised, the nuclease-defective RecBCD enzyme acts with the recF pathway to ensure

486 DNA repair by homologous recombination. In this model, we also propose that motor activity of

487 RecBCD enzyme is essential for rescuing P. syringae cells from low temperature induced

488 replication fork arrest.

489

490

21 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

491 MATERIALS AND METHODS 492 493 Bacterial strains, plasmids and growth conditions

494 The bacterial strains and plasmids used in this study are listed in Tables ST1. The

495 psychrophilic P. syringae Lz4W was isolated from a soil sample of Schirmacher Oasis, Antarctica

496 [40] and routinely grown at 22 or 4°C (for high and low temperatures respectively) in Antarctic

497 bacterial medium (ABM) composed of 5 g peptone and 2.0 g yeast extract per liter, as described

498 earlier [16]. E. coli strains were cultured at 37°C in Luria–Bertani (LB) medium, which contained

499 10 g tryptone, 5 g yeast extract and 10 g NaCl per liter. For solid media, 15 g bacto-agar (Hi-

500 Media) per liter was added to ABM or LB. When necessary, LB medium was supplemented with

501 ampicillin (100 μg ml−1), kanamycin (50 μg ml−1), gentamicin (15 μg ml−1) or tetracycline (20 μg

502 ml−1) for E. coli. For P. syringae, the ABM was supplemented with tetracycline (20 μg ml−1),

503 kanamycin (50 μg ml−1) as needed.

504 pBR322 plasmid DNA (4.3 Kb) and c+-3F3H dsDNA (a pBR322 derivative, containing two

505 sets of three tandem repeats of c sequences [27]) were purified using a Qiagen midi kit. Plasmids

506 were linearized with NdeI restriction endonuclease, dephosphorylated with SAP. The

507 dephosphorylated linear dsDNA was 5¢- labeled with T4-PNK and [g-32P] ATP as per the

508 manufacturer guidelines. DNA concentrations were determined by absorbance at 260 nm using

509 molar extinction co-efficient of 6500 M-1 cm-1 (in nucleotides). All restriction enzymes, DNA ligase

510 T4 Polynucleotide Kinase (T4 PNK), Shrimp alkaline phosphatase (SAP) and E. coli SSB were

511 purchased from New England Biolabs (MA, USA). Accuprime Pfx DNA polymerase was

512 purchased from Novagen (WI, USA).

513 Antibodies and Western analysis

514 Production of anti-RecB, anti-RecC and anti-RecD antibodies has been described [14].

515 For Western analysis, proteins were separated by SDS-PAGE, transferred onto Hybond C

516 membrane (Amersham Biosciences), and probed with appropriate antibodies. The

22 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

517 immunoreactive protein bands were detected by alkaline phosphatase-conjugated anti-rabbit goat

518 antibodies (Bangalore Genie, India). For quantification, the blots were scanned with a HP scanjet

519 and band intensities were measured using Image J software (rsbweb.nih.gov/ij/).

520 Over expression and purification of recombinant proteins

521 The LCBD (∆recBCD) strain harboring pGHCBD, pGHCBK28QD pGHCBDK229Q

522 pGHCBD1118AD plasmids [14] were initially grown in 10 ml ABM broth containing kanamycin (50

523 µg/ml) for 24 hrs at 22°C. 1% of above culture was inoculated into a 2-liter conical flask containing

524 500 ml ABM broth with kanamycin (50 µg/ml). The culture was then incubated at 22°C with

525 aeration for 24 hrs. Later, the bacterial cells were harvested by centrifugation at 4°C, 6000 rpm

526 for 10 min. The bacterial cell pellets were stored at -80°C. The cells pellet was removed as and

527 when required for the purification of proteins.

528 All the recombinant proteins in this study were expressed with His-tag on N-terminus of

529 RecC and purified by nickel-nitrilotriacetic acid (Ni-NTA) affinity chromatography as described in

530 the manufacturer’s protocol (Qiagen, New Delhi, India). In brief, cell lysate of overexpressed

531 strains containing His-tagged proteins were prepared by dissolving the cell pellet in 10 ml lysis

532 buffer (50 mM NaH2PO4 (pH 7.4), 300 mM NaCl, 10 mM imidazole and 10% glycerol) and lysed

533 by mild sonication. The sonicated cell lysate was then centrifuged with 14,000 rpm at 4°C, for 30

534 min to remove insoluble cellular debris. The supernatant was passed through a pre-equilibrated

535 column containing 1 ml slurry of Ni-NTA agarose beads and column was allowed to bind Ni-NTA

536 agarose beads with His-tagged proteins. Further, column was washed with 4-5 volumes of wash

537 buffer (50 mM NaH2PO4 (pH 7.4), 300 mM NaCl, 20 mM imidazole, and 10% glycerol). Finally,

538 the bound proteins were eluted with 1-2 ml elution buffer (50 mM NaH2PO4 (pH 7.4), / 300 mM

539 NaCl / 300 mM imidazole / 10% glycerol).

540 Gel filtration technique (size-exclusion column chromatography) was employed for the

541 purification of His-tagged RecBCD complex and the mutant protein complexes. For this,

23 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

542 Superose-12 gel filtration column (Pharmacia Fine Chemicals) was used in the fast protein liquid

543 chromatography (FPLC) (Pharmacia Fine Chemicals). Initially, the column was pre-equilibrated

544 with buffer contained 20 mM Tris HCl (pH 7.5), 0.1 mM EDTA, 150 mM NaCl, 0.1 mM PMSF and

545 10% glycerol. Later, 0.5 ml of Ni-NTA purified protein solution was injected to the column and

546 allowed to pass through the column at a flow rate of 0.4 ml/min. Optical density at 280 nm was

547 recorded for the eluted protein fractions and the fractions were collected in separate

548 microcentifuge tubes. The protein fractions were then analyzed on SDS-PAGE stained with

549 coomassie brilliant blue or silver nitrate. The gel filtration protein fractions of interest were further

550 membrane dialyzed in 50% glycerol containing gel filtration buffer. RecBCD enzyme

551 concentrations were determined by measuring OD280 and using molar extinction co-efficient 4.7 X

552 105 M-1 cm-1 as determined by ExPASy – ProtParam tool (https://web.expasy.org/protparam).

553 Thin-layer chromatography based assay

554 ATPase activity of RecBCD and mutant proteins was assayed at different temperatures

555 by thin layer chromatography (TLC) on polyethylene-imine (PEI)-cellulose plates (E-Merck,

556 Germany). The assay was performed as described earlier [41,42]. Assays were carried out at

557 indicated temperatures in a reaction volume of 20 µl containing 25 mM Tris acetate (pH 7.5), 1

558 mM Mg acetate, 1 mM DTT, 100mM NdeI linearized pBR322-dsDNA and 200 µM ATP as a

559 substrate with 2 nM RecBCD and mutant enzymes. One µl of 100 times diluted 10 mCi ml-1 stock

560 of [g-32P] ATP (specific activity 3000 Ci mmol-1) was used as a tracer in each reaction to measure

561 the rate of ATP hydrolysis. Following 0, 1, 2, 3, 5 10 minutes of reaction, 0.5 µl aliquots of the

562 samples were spotted on TLC plate, air-dried and were allowed to develop in a mobile phase

563 containing 0.5 M formic acid and 0.5 M lithium chloride for 15 minutes. The TLC plate was dried

564 and exposed to the Phosphor imaging plate for 4-6 hrs. The Imaging plates were scanned in a

565 Phosphor Imager, and the amounts of 32Pi and [g-32P] ATP were quantified using Image gauge

566 software (Fuji-3000). Further, data were analyzed using GraphPad Prism 4.0 software.

24 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

567 DNA unwinding assay

568 Plasmid DNAs were linearized with appropriate restriction enzymes in the presence of

569 shrimp alkaline phosphatase and 5`-end-labeled by T4 polynucleotide kinase and [g-32P] ATP.

570 Subsequent purification of labeled DNA was accomplished by passage through a MicroSpin S-

571 200 HR column (Amersham biosciences-GE healthcare, Buckinghamshire, UK). The reaction

572 mixtures contained 25 mM Tris acetate (pH 7.5), 2 mM magnesium acetate (as indicated), 1 mM

573 DTT, 10 µM (nucleotides) linear pBR322 dsDNA P32-labeled at 5′- end, 5 mM ATP, 2 µM E. coli

574 SSB protein and 0.5 nM RecBCDPs or mutant enzymes. DNA unwinding reactions were started

575 with the addition of either enzyme or ATP, after pre-incubation of all other components at 22 or

576 4°C for 5 min. Assays were stopped at the indicated times by addition of proteinase K to a final

577 concentration of 0.5 mg/ml, which was dissolved in sample loading buffer (125 mM EDTA, 40%

578 glycerol, 2.5% SDS, 0.25% bromophenol blue, and 0.25% xylene cyanol). After 5-min incubation

579 with proteinase K at room temperature, the reaction products were run on a 1% (w/v) agarose gel

580 in a 1X TBE (45 mM Tris borate (pH 8.3) and 1 mM EDTA) buffer at 25-30 constant volts for 15

581 hrs. Agarose gels were dried, exposed to phosphor imaging plates and quantified using Phosphor

582 Imager (Fuji-3000). Further, data were analyzed using image gauge software.

583 The DNA unwinding rates of were measured by using the following formula

-1 -1 584 nM bp s (nM helicase) = slope X (C/ [100%]) X ts X (1/E)

585 Where C is the concentration of linear dsDNA substrate in base-pairs in nM (i.e., 5000 nM), ts is

586 the time in seconds, and E is the enzyme concentration in nM.

587 DNA degradation assay

588 The assays were performed as described above in DNA unwinding assay, except that the

589 reaction mixtures contained 6 mM magnesium acetate and 2 mM ATP.

590 Single-stranded DNA endonuclease assay

25 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

591 The endonuclease activity of RecBCD enzyme on ssDNA was examined using a circular

592 M13 ssDNA substrate as described previously [20,22]. In brief, endonuclease activity was tested

593 in 3 different buffer conditions. The first reaction mixture contained 50 mM MOPS (pH=7.5), 1 mM

594 ATP, 10 mM MgCl2, 4.16 nM circular M13 ssDNA with 0.5 nM RecBCD. The second reaction

595 mixture contained 25 mM Tris-acetate, 1 mM ATP, 8 mM Mg-acetate,1 mM DTT, 4.16 nM M13

596 ssDNA with 0.5 nM RecBCD. The third reaction mixture contained 25 mM Tris-acetate, 2 mM

597 ATP, 6 mM Mg-acetate, 1 mM DTT, 4.16 nM M13 ssDNA with 0.5 nM RecBCD. After the reaction,

598 samples were removed at the indicated times; quenched with 120 mM EDTA, 40% (v/v) glycerol,

599 and 0.125% bromphenol blue; and loaded on a 0.8% agarose gel in 1X TBE (90 mM Tris borate,

600 2 mM EDTA). The gel was run at 4 V/cm for 3 h and stained with ethidium bromide (0.5 mg/ml).

601 The bands were visualized by exposure to UV light.

602 Site directed deletion of regions from plasmid pBR322

603 Different internal regions of plasmid pBR322 were deleted using site directed deletion. In short,

604 primers were designed consists of sequences flanking the region to be deleted. PCR was

605 performed to amplify whole Plasmid DNA using Accuprime Pfx DNA polymerase (Invitrogen).

606 After PCR, reaction mixture was subjected for the overnight DpnI digestion. 5 μl of reaction

607 mixture was then transformed into DH5α ultra-competent cells. All selected regions, which had to

608 be deleted was in tetracycline resistance gene of the plasmid. Therefore, for primary screening

609 only those colonies were selected which were AmpR and TetS. Deletion was further confirmed by

610 PCR and sequencing. All primer list and corresponding deleted regions are shown in Table ST2.

611

612 ACCESSION NUMBERS

613 Accession no. AOGS01000000, the draft genome sequence of P. syringae Lz4W and,

614 Accession no. NC_007492.2, the genome sequence Pseudomonas fluorescens Pf0-1.

615

616 SUPPLIMENTARY MATERIALS

26 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

617 Supplementary material related to this manuscript is attached.

618 ACKNOWLEDGEMENTS

619 Research in M.K.R. laboratory is supported by the Council of Scientific and Industrial Research

620 (CSIR), Government of India. A part of the work was supported by a grant to M.K.R. from

621 Department of Science and Technology (DST), Government of India. T.L.P and A.K.S.

622 acknowledge CSIR, India for research fellowships.

623 We also thank Prof. Stephen C Kowalczykowski, Prof. Michel Benedicte and Dr. Naofumi Handa

624 for their critical reading and comments on the manuscript.

625

626 CONFLICT OF INTEREST

627 The authors declare that they have no conflict of interest with the contents of this article.

628 AUTHOR CONTRIBUTIONS

629 Theetha L. Pavankumar & Anurag Kumar Sinha both have contributed equally to this work.

630 T.L.P., A.K.S. and M.K.R designed the experiments. T.L.P. and A.K.S. conducted the experiment.

631 T.L.P., A.K.S. and M.K.R analyzed the results and wrote the manuscript.

632

27 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

633 FIGURE LEGENDS

634 FIGURE 1. Purification of His-tagged RecBCD complex and mutant enzymes by Ni-NTA

635 agarose column chromatography. (A) Steps involved in purification of wild-type RecBCD and

636 mutant proteins. (B) SDS-PAGE analysis of purified RecBCD protein fractions of wild-type and

637 mutants. Purified protein fractions were stained either with coomassie brilliant blue (B) or with

638 silver nitrate (C). Three protein bands of expected size corresponding to RecB, RecC and RecD

639 are visible on the gel. A low molecular protein (~60 kDa) observed on silver nitrate strained gel

640 was identified as GroEL, a HSP-60 family chaperonin.

641

642 FIGURE 2. ATP hydrolysis of wild-type and mutant RecBCD enzymes at different

643 temperatures. The assays were carried out by TLC method as described in Methods. (A) A graph

644 showing the concentration dependent ATP hydrolysis by RecBCD (WT) and mutant enzymes at

645 22°C. The inset in B shows a blow-up of the same data of mutant enzymes using an expanded y-

646 scale. (B) Representative of TLC plates showing the ATP hydrolysis by wild-type and mutant

647 RecBCD enzymes at 37°, 22° and 4°C. (C) ATPase activities at 37°, 22° and 4°C are shown in

648 histogram with error bars. The ATP hydrolysis data presented in the histogram represents the

649 results obtained from three independent experiments.

650

651 FIGURE 3. Effects of magnesium and ATP on the unwinding and degradation by wild-type

652 RecBCD enzyme of P. syringae. The DNA unwinding and degradation assays were performed

653 with [5ʹ-32P] labeled NdeI linearized pBR322 plasmid DNA in the presence of different

654 concentrations of magnesium and ATP, as described in Methods; (A) the reaction mixture

655 contained fixed amount of Mg-acetate (2 mM) and the amount ATP was varied (0 to 10 mM) as

656 indicated. (B) The reaction mixture contained the fixed amount of ATP (2 mM) and concentration

657 of Mg-acetate was varied (0 to 10 mM) as indicated. The reactions were performed for 5 min and

28 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

658 analyzed on a 1% agarose gel containing 1X TBE buffer at a 25-30 Volts for 15 hrs. Agarose gels

659 were dried, exposed to phosphor imaging plates and quantified using Phosphor Imager (Fuji-

32 660 3000). The data were analyzed using image gauge software. The lanes C1 and C2 contain [5ʹ- P]

661 labeled dsDNA and the heat denatured 5ʹ-32P labeled dsDNA respectively as a control. A [5ʹ-32P]

662 labeled discrete DNA bands of smaller than full-length ssDNA of pBR322 were also noticed in the

663 lanes 5-10 of panel B.

664

665 FIGURE 4. DNA unwinding and degradation of NdeI digested linear dsDNA of pBR322 by

666 RecBCDPs (WT) enzyme at 22° and 4°C. The DNA unwinding and degradation at 22° and 4°C

667 were carried out as described in Methods. (A) The DNA unwinding reactions contained 5 mM

668 ATP and 2 mM Mg++ (limiting magnesium condition) (B) The DNA degradation reactions contained

669 2 mM ATP and 6 mM Mg++ (excess magnesium condition). The reactions were initiated by adding

670 ATP, and stopped at the indicated times by adding stop-buffer. The lanes C1 and C2 contain [5ʹ-

671 32P] labeled NdeI linearised double-stranded and the heat-denatured ssDNA of pBR322 as

672 control. The discrete ssDNA fragments of pBR322 produced by the nuclease activity of RecBCD

673 enzyme are also indicated.

674

675 FIGURE 5. DNA unwinding and degradation of NdeI linearized dsDNA of pBR322 by mutant

676 RecBCDPs enzymes at 22° and 4°C. (A) DNA unwinding and degradation by RecBK28QCD

677 enzyme. The DNA unwinding and degradation at 22° and 4°C were carried out as described in

678 Methods. Note that this enzyme shows DNA unwinding (ssDNA production), but no detectable

679 DNA degradation at 22°C, and none were detectable at 4°C. (B) DNA unwinding and

680 degradation by RecBCDK229Q enzyme. The RecBCDK229Q enzyme has apparently retained both

681 the DNA unwinding and degradation properties. But, interestingly, the discrete DNA bands are

682 absent (C) DNA unwinding by nuclease-deficient RecBD1118ACD enzyme under limiting and

29 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

683 excess magnesium conditions at 22 and 4°C. RecBD1118ACD enzyme is unable to degrade

684 DNA at both 22 and 4°C and notably, the DNA unwinding seems to be faster in excess-

685 magnesium reaction condition (2 mM ATP:6 mM Mg++) compared to limiting magnesium reaction

686 condition (5 mM ATP:2 mM Mg++). Each reaction mixtures contained 0.5 nM of enzyme and 10

32 32 687 µM (nucleotides) linear [5ʹ- P] labeled pBR322 dsDNA. The lanes C1 and C2 contain [5ʹ- P]

688 labeled NdeI linearized double-stranded and the heat-denatured ssDNA of pBR322 respectively.

689

690 FIGURE 6. Identification of Chi sequence using PCR amplified pBR322 fragment (3.6 kb)

691 containing internal deletions as a DNA substrate. (A) Top panel; Schematic representation of

692 NdeI linearized pBR322 plasmid DNA. The locations of OROPI and OPROPII primers used in

693 PCR amplification of 3.6 kb fragments are indicated. Bottom panel; DNA degradation reactions

694 performed using the PCR amplified 3.6 kb DNA as substrate, having either bottom strand labeled

695 (left panel) or top strand labeled (middle panel) or both strand labeled (right panel). The lane C

696 contains [5ʹ-32P] labeled heat-denatured ssDNA as a control. Notably, the top strand labeling of

697 DNA substrate resulted in appearance of discrete DNA fragments. (B) Deletion of 400-450 bp of

698 pBR322 (pBR322(∆400-450)) resulted in disappearance of DNA fragments. But, it is clearly visible

699 when intact 3.6 kb pBR322 was used as a substrate, suggesting the presence of putative ChiPs

700 sequence in this region. (C) Further deletion of 401-419; 421-439 bo regions; and, (D) 421-429;

701 431-439; 441-449 bp regions of pBR322 shows that the DNA fragments are indeed from the 431-

702 439 bp region of the pBR322 DNA sequence. (E) A schematic representation of 3.6 kb region of

703 pBR322 and deleted regions within, are shown in the left panel. Right panel shows the presence

704 (+) or absence (-) of intense protected band (Chi-like fragments), when these constructs are used

705 as assay substrates.

706

30 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

707 FIGURE 7. (A) DNA degradation of modified fragment of pBR322 by RecBCDPs enzyme.

708 Apart from one 8-mer ChiPs sequence at 431th position of pBR322, there are two 7-mer similar

709 sequences present in the plasmid pBR322. One such similar sequence at 964th position (5′

710 GCTGGCGT 3′) was mutated to make it an 8-mer sequence (5′ GCTGGCGC 3′). This substrate

711 (pBR322T971C) that has two Chi sequences in the same orientation when used as a substrate,

712 yielded two DNA fragments with high intensity. (B) The DNA degradation pattern of XbaI

713 digested linear dsDNA of pBKS plasmid by RecBCDPs enzyme. ChiPs are inserted into

714 pBluescript vector (pBKS) by site directed insertion. XbaI linearized pBKS vector or pBKS with

715 ChiPs was 5ʹ-end labeled with 32P has been used for assays. The discrete ssDNA of expected size

716 was produced when pBKS (ChiPs) was used as a substrate, whereas it is absent, when native

717 pBKS was used.

718

719 FIGURE 8. Role of RecBCD dependent DNA repair pathway in rescue of low temperature

720 induced replication forks arrest. (i) A chromosome replicating bi-directionally, (ii) encounters

721 low temperature-induced chromosomal lesion or blockage, (iii), causing replication fork arrest

722 and fork reversal (RFR). RFR is suppressed by linearized chromosomal DNA degradation by

723 RecBCD enzyme and resetting of replication fork. (iv) RFR is stabilized further by RuvAB

724 complex and, (v) further, resolved by RuvC leading to chromosomal linearization. Linearized

725 chromosome is processed by either RecBCD alone or in conjunction (when RecBCD is

726 nuclease defective) with recFOR-recJ hybrid pathway. The defective motors activity of RecBCD

727 enzyme leads to chromosomal fragmentation and cell death at the low temperature.

728

31 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

729 REFERENCES

730 1. Cromie, G.A.; Connelly, J.C.; Leach, D.R. Recombination at double-strand breaks and 731 DNA ends: Conserved mechanisms from phage to humans. Mol Cell 2001, 8, 1163- 732 1174. 733 2. Dillingham, M.S.; Kowalczykowski, S.C. Recbcd enzyme and the repair of double- 734 stranded DNA breaks. Microbiol Mol Biol Rev 2008, 72, 642-671, Table of Contents. 735 3. Kowalczykowski, S.C.; Dixon, D.A.; Eggleston, A.K.; Lauder, S.D.; Rehrauer, W.M. 736 Biochemistry of homologous recombination in escherichia coli. Microbiol Rev 1994, 737 58, 401-465. 738 4. Michel, B.; Boubakri, H.; Baharoglu, Z.; LeMasson, M.; Lestini, R. Recombination 739 proteins and rescue of arrested replication forks. DNA Repair (Amst) 2007, 6, 967- 740 980. 741 5. Kuzminov, A. Single-strand interruptions in replicating chromosomes cause double- 742 strand breaks. Proc Natl Acad Sci U S A 2001, 98, 8241-8246. 743 6. Taylor, A.F.; Smith, G.R. Strand specificity of nicking of DNA at chi sites by recbcd 744 enzyme. Modulation by atp and magnesium levels. J Biol Chem 1995, 270, 24459- 745 24467. 746 7. Stahl, F.W.; Stahl, M.M. Recombination pathway specificity of chi. Genetics 1977, 86, 747 715-725. 748 8. Anderson, D.G.; Kowalczykowski, S.C. The recombination hot spot chi is a regulatory 749 element that switches the polarity of DNA degradation by the recbcd enzyme. Genes 750 Dev 1997, 11, 571-581. 751 9. Anderson, D.G.; Kowalczykowski, S.C. The translocating recbcd enzyme stimulates 752 recombination by directing reca protein onto ssdna in a chi-regulated manner. Cell 753 1997, 90, 77-86. 754 10. Hiom, K.; West, S.C. Branch migration during homologous recombination: Assembly 755 of a ruvab-holliday junction complex in vitro. Cell 1995, 80, 787-793. 756 11. Pavankumar, T.L.; Exell, J.C.; Kowalczykowski, S.C. Direct fluorescent imaging of 757 translocation and unwinding by individual DNA helicases. Methods Enzymol 2016, 758 581, 1-32. 759 12. Spies, M.; Bianco, P.R.; Dillingham, M.S.; Handa, N.; Baskin, R.J.; Kowalczykowski, S.C. 760 A molecular throttle: The recombination hotspot chi controls DNA translocation by 761 the recbcd helicase. Cell 2003, 114, 647-654. 762 13. Spies, M.; Amitani, I.; Baskin, R.J.; Kowalczykowski, S.C. Recbcd enzyme switches 763 lead motor subunits in response to chi recognition. Cell 2007, 131, 694-705. 764 14. Pavankumar, T.L.; Sinha, A.K.; Ray, M.K. All three subunits of recbcd enzyme are 765 essential for DNA repair and low-temperature growth in the antarctic pseudomonas 766 syringae lz4w. PLoS One 2010, 5, e9412. 767 15. Sinha, A.K.; Pavankumar, T.L.; Kamisetty, S.; Mittal, P.; Ray, M.K. Replication arrest is 768 a major threat to growth at low temperature in antarctic pseudomonas syringae 769 lz4w. Mol Microbiol 2013, 89, 792-810. 770 16. Regha, K.; Satapathy, A.K.; Ray, M.K. Recd plays an essential function during growth 771 at low temperature in the antarctic bacterium pseudomonas syringae lz4w. Genetics 772 2005, 170, 1473-1484.

32 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

773 17. Handa, N.; Bianco, P.R.; Baskin, R.J.; Kowalczykowski, S.C. Direct visualization of 774 recbcd movement reveals cotranslocation of the recd motor after chi recognition. 775 Mol Cell 2005, 17, 745-750. 776 18. Dixon, D.A.; Kowalczykowski, S.C. Role of the escherichia coli recombination hotspot, 777 chi, in recabcd-dependent homologous pairing. J Biol Chem 1995, 270, 16360- 778 16370. 779 19. Rosamond, J.; Telander, K.M.; Linn, S. Modulation of the action of the recbc enzyme 780 of escherichia coli k-12 by ca2+. J Biol Chem 1979, 254, 8646-8652. 781 20. Eggleston, A.K.; Kowalczykowski, S.C. Biochemical characterization of a mutant 782 recbcd enzyme, the recb2109cd enzyme, which lacks chi-specific, but not non- 783 specific, nuclease activity. J Mol Biol 1993, 231, 605-620. 784 21. Sun, J.Z.; Julin, D.A.; Hu, J.S. The nuclease domain of the escherichia coli recbcd 785 enzyme catalyzes degradation of linear and circular single-stranded and double- 786 stranded DNA. Biochemistry 2006, 45, 131-140. 787 22. Yu, M.; Souaya, J.; Julin, D.A. The 30-kda c-terminal domain of the recb protein is 788 critical for the nuclease activity, but not the helicase activity, of the recbcd enzyme 789 from escherichia coli. Proc Natl Acad Sci U S A 1998, 95, 981-986. 790 23. Dixon, D.A.; Kowalczykowski, S.C. The recombination hotspot chi is a regulatory 791 sequence that acts by attenuating the nuclease activity of the e. Coli recbcd enzyme. 792 Cell 1993, 73, 87-96. 793 24. McKittrick, N.H.; Smith, G.R. Activation of chi recombinational hotspots by recbcd- 794 like enzymes from enteric bacteria. J Mol Biol 1989, 210, 485-495. 795 25. Lehman, I.R.; Nussbaum, A.L. The deoxyribonucleases of escherichia coli. V. On the 796 specificity of exonuclease i (phosphodiesterase). J Biol Chem 1964, 239, 2628-2636. 797 26. Dixon, D.A.; Churchill, J.J.; Kowalczykowski, S.C. Reversible inactivation of the 798 escherichia coli recbcd enzyme by the recombination hotspot chi in vitro: Evidence 799 for functional inactivation or loss of the recd subunit. Proc Natl Acad Sci U S A 1994, 800 91, 2980-2984. 801 27. Anderson, D.G.; Churchill, J.J.; Kowalczykowski, S.C. A single mutation, recb(d1080a,) 802 eliminates reca protein loading but not chi recognition by recbcd enzyme. J Biol 803 Chem 1999, 274, 27139-27144. 804 28. Taylor, A.F.; Schultz, D.W.; Ponticelli, A.S.; Smith, G.R. Recbc enzyme nicking at chi 805 sites during DNA unwinding: Location and orientation-dependence of the cutting. 806 Cell 1985, 41, 153-163. 807 29. Amundsen, S.K.; Taylor, A.F.; Smith, G.R. The recd subunit of the escherichia coli 808 recbcd enzyme inhibits reca loading, homologous recombination, and DNA repair. 809 Proc. Natl. Acad. Sci. U. S. A. 2000, 97, 7399-7404. 810 30. Churchill, J.J.; Anderson, D.G.; Kowalczykowski, S.C. The recbc enzyme loads reca 811 protein onto ssdna asymmetrically and independently of chi, resulting in 812 constitutive recombination activation. Genes Dev 1999, 13, 901-911. 813 31. Spies, M.; Dillingham, M.S.; Kowalczykowski, S.C. Translocation by the recb motor is 814 an absolute requirement for {chi}-recognition and reca protein loading by recbcd 815 enzyme. J Biol Chem 2005, 280, 37078-37087. 816 32. Ivancic-Bace, I.; Peharec, P.; Moslavac, S.; Skrobot, N.; Salaj-Smic, E.; Brcic-Kostic, K. 817 Recfor function is required for DNA repair and recombination in a reca loading- 818 deficient recb mutant of escherichia coli. Genetics 2003, 163, 485-494.

33 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

819 33. Miranda, A.; Kuzminov, A. Chromosomal lesion suppression and removal in 820 escherichia coli via linear DNA degradation. Genetics 2003, 163, 1255-1271. 821 34. El Karoui, M.; Biaudet, V.; Schbath, S.; Gruss, A. Characteristics of chi distribution on 822 different bacterial genomes. Res Microbiol 1999, 150, 579-587. 823 35. Arnold, D.A.; Bianco, P.R.; Kowalczykowski, S.C. The reduced levels of chi recognition 824 exhibited by the recbc1004d enzyme reflect its recombination defect in vivo. J Biol 825 Chem 1998, 273, 16476-16486. 826 36. Handa, N.; Yang, L.; Dillingham, M.S.; Kobayashi, I.; Wigley, D.B.; Kowalczykowski, 827 S.C. Molecular determinants responsible for recognition of the single-stranded DNA 828 regulatory sequence, chi, by recbcd enzyme. Proc Natl Acad Sci U S A 2012, 109, 829 8901-8906. 830 37. Arakawa, K.; Uno, R.; Nakayama, Y.; Tomita, M. Validating the significance of 831 genomic properties of chi sites from the distribution of all octamers in escherichia 832 coli. Gene 2007, 392, 239-246. 833 38. Winsor, G.L.; Griffiths, E.J.; Lo, R.; Dhillon, B.K.; Shay, J.A.; Brinkman, F.S. Enhanced 834 annotations and features for comparing thousands of pseudomonas genomes in the 835 pseudomonas genome database. Nucleic Acids Res 2016, 44, D646-653. 836 39. Lovett, S.T.; Luisi-DeLuca, C.; Kolodner, R.D. The genetic dependence of 837 recombination in recd mutants of escherichia coli. Genetics 1988, 120, 37-45. 838 40. Shivaji, S.; Rao, N.S.; Saisree, L.; Sheth, V.; Reddy, G.S.; Bhargava, P.M. Isolation and 839 identification of pseudomonas spp. From schirmacher oasis, antarctica. Appl Environ 840 Microbiol 1989, 55, 767-770. 841 41. Korangy, F.; Julin, D.A. Alteration by site-directed mutagenesis of the conserved 842 lysine residue in the atp-binding consensus sequence of the recd subunit of the 843 escherichia coli recbcd enzyme. J Biol Chem 1992, 267, 1727-1732. 844 42. Kornberg, A.; Scott, J.F.; Bertsch, L.L. Atp utilization by rep protein in the catalytic 845 separation of DNA strands at a replicating fork. J Biol Chem 1978, 253, 3298-3304. 846 43. Smith, G.R.; Comb, M.; Schultz, D.W.; Daniels, D.L.; Blattner, F.R. Nucleotide sequence 847 of the chi recombinational hot spot chi +d in bacteriophage lambda. J Virol 1981, 37, 848 336-342. 849 44. Chedin, F.; Noirot, P.; Biaudet, V.; Ehrlich, S.D. A five-nucleotide sequence protects 850 DNA from exonucleolytic degradation by addab, the recbcd analogue of bacillus 851 subtilis. Mol Microbiol 1998, 29, 1369-1377. 852 45. Halpern, D.; Chiapello, H.; Schbath, S.; Robin, S.; Hennequet-Antier, C.; Gruss, A.; El 853 Karoui, M. Identification of DNA motifs implicated in maintenance of bacterial core 854 genomes by predictive modeling. PLoS Genet 2007, 3, 1614-1621. 855

34 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

856 TABLE 1. ATPase activity of wild type and mutant RecBCD enzymes

° ° ° a a Enzyme 37 C 22 C 4 C Km kcat

RecBCD (WT) 366.7 ± 30.9 236.5 ± 9.4 119.0 ± 19.6 57.8 ± 10.1 61.3

RecBD1118ACD 346.0 ± 33.8 209.8 ± 13.9 112.4 ± 23.3 59.5 ± 8.5 55.3

RecBK28QCD 33.3 ± 0.9 16.2 ± 3.0 9.7 ± 1.5 58.3 ± 11.0 6.6

RecBCDK229Q 30.5 ± 1.4 26.9 ± 2.2 9.7 ± 1.7 49.4 ± 5.8 6.1

857 ATPase activities are expressed here in units of µmol ATP hydrolyzed per second per µmol

a 858 RecBCD enzymes. Km and kcat values are calculated from ATPase activity at 22 °C.

859

860 TABLE 2. DNA unwinding and degradation activities of the wild type and mutant RecBCD

861 enzymes at 22 and 4°C

Rate of DNA Rate of DNA Rate of DNA Rate of DNA unwinding a unwinding a or Enzyme unwinding unwinding or degradation degradation at 22°C (bp s-1) at 4°C (bp s-1) at 22°C (bp s-1) at 4°C (bp s-1) ATP > Mg++ ATP > Mg++ ATP < Mg++ ATP < Mg++

(5:2 mM) (5:2 mM) (2:6 mM) (2:6 mM)

RecBCDPs 35.1 ± 1.6 ND b 101.5 ± 3.3 55.8 ± 6.9

RecBD1118ACD 30.4 ± 1.7 ND 109.7 ± 17.5 41.0 ± 8.1

RecBK28QCD 1.5 ± 0.76 ND 12.9 ± 3.4 ND

RecBCDK229Q 27.2 ± 6.1 ND 92.6 ± 14.1 17.3 ± 9.6

862 The values shown for the DNA unwinding/degradation rates with mean ± S.E. from two

863 independent experiments. a Rate of DNA unwinding under excess magnesium reaction condition.

864 b ND - not determined

865

35 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

866 TABLE 3. Chi sequences identified in different bacteria.

Bacteria Chi sequence Recombination machinery Escherichia coli 5′ GCTGGTGG 3′ RecBCD [43] Bacillus subtilis 5′ AGCGG 3′ AddAB [44] Lactococcus lactis 5′ GCGCGTG 3′ RexAB [34] Streptococcus pneumoniae 5′ GCGCGTG 3′ RexAB [45] Staphylococcus aureus 5′ GAAGCGG 3′ - [45] Streptococcus agalactiae 5′ GCGCGTG 3′ - [45] Streptococcus thermophilus 5′ GCGCGTG 3′ - [45] Pseudomonas syringae Lz4W 5′ GCTGGCGC 3′ RecBCD [14] 867 868

36 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

A B C

Overexpression of RecBCD and mutant proteins in LCBD strain (!recCBD::Tet) RecB RecB His-RecC His-RecC Cell lysis by sonication RecD RecD Ni-NTA- affinity column purification GroEL

Superose-12 column gel filtration

Figure 1 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

A 250

225

200

) WT

-1 RecBCD 175 RecBD1118ACD 150

30 125 molprotein s 25 µ 100 20

ATPaseactivity 15 K28Q molper 75 RecB CD 10 µ ( K229Q 5 50 RecBCD 0 0 200 400 600 800 1000 25

0 0 200 400 600 800 1000 1200 ATP (µM) B RecBHis-CD (WT) RecBD1118A His-CD 37°C 22°C 4°C 37°C 22°C 4°C 0 2 5 0 2 5 0 2 5 0 2 5 0 2 5 0 2 5 min !P-32

ATP-!P-32

RecBK28Q His-CD RecBHis-CDK229Q 37°C 22°C 4°C 37°C 22°C 4°C 0 2 5 0 2 5 0 2 5 0 2 5 0 2 5 0 2 5 min !P-32

ATP-!P-32

C

400 ) -1

300 37°C 22°C 4°C 200 molenzyme s µ

ATP hydrolysis ATP 100 molper µ ( 0 CD CD K229Q D1118A K28Q RecBCD (WT)RecB RecB RecBCD

Figure 2 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

A

ATP - - 0 1 2 3 4 5 6 7 8 10 mM Mg2+ - - 2 2 2 2 2 2 2 2 2 2 mM

Lane # C1 C2 1 2 3 4 5 6 7 8 9 10

B Mg2+ - - 0 1 2 3 4 5 6 7 8 10 mM ATP - - 2 2 2 2 2 2 2 2 2 2 mM

Lane # C1 C2 1 2 3 4 5 6 7 8 9 10

Discrete DNA fragments

Figure 3 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

A Limiting magnesium condition ATP (5 mM) : Mg++ (2 mM) 22°C 4°C

0 1 2 5 10 C1 C2 0 1 2 5 10 min

B

Excess magnesium condition ATP (2 mM) : Mg++ (6 mM) 22°C 4°C

0 1 2 5 10 C1 C2 0 1 2 5 10 min

Discrete DNA fragments

Figure 4 bioRxiv preprint doi: https://doi.org/10.1101/266460; this version posted February 15, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. Limiting A Excess magnesium RecBK28QCD magnesium 22°C 22°C 4°C

C1 C2 0 1 2 5 10 0 1 2 5 10 0 1 2 5 10 min

B Limiting Excess magnesium RecBCDK229Q magnesium 22°C 22°C 4°C

0 1 2 5 10 C1 C2 0 1 2 5 10 0 1 2 5 10 min

Limiting C Excess magnesium RecBD1118ACD magnesium 22°C 22°C 4°C

C1 C2 0 1 2 5 0 1 2 5 0 1 2 5 10 min

Figure 5 OROPI OROPII 3650 bp pBR322 5’ A E NdeI (2296) EcoRI (4361/1) rop region pBR322 ( full length-4361 bp) Chi fragment/s NdeI (2296) (4361/1) C 0 1 3 5 0 1 3 5 0 1 3 5 min 5’ 3’ + 3.6 kb fragment of pBR322 (4361/1) OROPII (2351) 5’ Discrete 3’ + DNA OROPI (1641) fragments

Deletion of 273-381 region from 3.6 kb fragment (4361/1) OROPII (2351) (!273-381) 5’ + 3’ OROPI (1641)

Deletion of 400-451 region from 3.6 kb fragment (4361/1) OROPII (2351) (!400-451) 5’ - 3’ B OROPI (1641) pBR322 pBR322(!400-450) C 0 1 2 3 5 C 0 1 2 3 5 min 400 450 !400 419 + !421 439 - !421 429 + prominent !431 439 (5’- GCTGGCGC -3’) - DNA !441 449 fragment +

C pBR322 pBR322(!421-439) pBR322(!401-419) C 0 1 2 3 5 C 0 1 2 3 5 C 0 1 2 3 5 min

prominent DNA fragment

D pBR322(!421-429) pBR322(!431-439) pBR322(!441-449) C 0 1 2 3 5 C 0 1 2 3 5 C 0 1 2 3 5 min

prominent DNA fragment

Figure 6 A

pBR322 pBR322T971C C 0 5 0 5 min

dsDNA (3650 bp) ✸ ✸

✸ + ✸ prominent ssDNA DNA fragments

B !Ps 2173 bp 786 bp pBKS-XbaI pBKS (ChiPs)-XbaI 0 1 2 3 5 C 0 1 2 3 5 min

dsDNA + DNA ssDNA fragment

Figure 7 Low temperature induced DNA damage and replication arrest Replicating chromosome (i) (ii) (iii)

RF reversal (RFR) Linear chromosomal DNA degradation by RecBCD enzyme and Replication fork resetting RecBD1118ACD (helicase), RecJ-RecFOR hybrid pathway RecBCD pathway

Defective motor activity of RecBCD enzyme

(iv) (v) (vi) HJ resolution and RecBCD dependent linearization of degradation of RFR Chromosome fragmentation chromosome Cell death and cold sensitivity RecBCD RuvAB RuvC

Figure 8